Published on 21 September 2026
Intermittent theta burst stimulation for depressed adolescents: what does the evidence actually show?
Journal of Affective Disorders · 2026 · Deng et al.
DOI 10.1016/j.jad.2025.120435
PMID 41101467
Scientific 73
Editorial 78
The essentials
Five randomized trials, 264 adolescents, six active arms, a weighted mean age of fifteen and a half, all of them conducted in China. The review itself is carefully done: searches run in both English and Chinese across six databases, quality assessed independently by two tools, and a conclusion that oversells nothing. The evidence is not sufficient to establish a clinically meaningful effect of intermittent theta burst stimulation in depressed adolescents, whether delivered in its standard daily form or in an accelerated schedule. The authors state that a meta-analysis was not feasible given the amount of data available. One isolated signal deserves to be followed: a reduction in suicidal ideation under the daily protocol, reported in a single trial, the very one the authors judge probably among the most exposed to bias. Two findings call for caution: adverse events were systematically recorded in only one of the five trials, and one of the review’s two headline conclusions is that both protocols were associated with a higher incidence of headache.
Context
Repetitive transcranial magnetic stimulation is now well established in adult depression, and intermittent theta burst stimulation gained traction because it shortens session length considerably. Accelerated protocols, which condense at least two sessions a day, have since gained ground, with the authors noting how quickly Stanford’s neuromodulation protocol became known in adults.
The demand is now reaching child and adolescent psychiatry, driven by families who have read something about it and by the genuine difficulty of treating adolescent depression. In practice, the question parents ask is simple: does it work, and is it safe. This review allows an honest answer to both, and it is not the answer anyone was hoping for.
The study at a glance
| Population | |
| Adolescents with a major depressive episode or bipolar depression, diagnosed by DSM-IV-TR or DSM-5 criteria. Weighted mean age fifteen and a half, males 41.0% of the sample. 264 participants in total, across five trials and six active arms: 54 on the daily protocol, 96 on the accelerated protocol, 114 on sham stimulation. Two trials included participants up to 24 and 25 years old, extending beyond strict adolescence. | |
| Intervention | |
| Intermittent theta burst stimulation, in a daily or an accelerated protocol, added to ongoing treatment. | |
| Comparator | |
| Sham stimulation. | |
| Outcomes | |
| Depressive symptoms measured with standard scales, anxiety symptoms, suicidal ideation, cognitive function, trial dropout, and adverse events. | |
| Design | |
| Systematic review of randomized trials reported per PRISMA, search conducted through February 17, 2025 across six databases, four international (PubMed, EMBASE, PsycINFO, Cochrane Library) and two Chinese (WanFang, CNKI). Quality assessed with the Jadad scale and the Cochrane risk-of-bias tool. Narrative synthesis with no quantitative pooling: the authors state that meta-analysis was not feasible given the amount of data available, and note substantial heterogeneity across interventions. The source does not report an Oxford level of evidence. |
Quality control
| Item checked | Verdict |
|---|---|
| Conduct of the review | Rigorous |
| FindingSearches were run in both English and Chinese across six databases, widening the capture net. Still, all five included trials were published in English-language international journals: the Chinese-language search did not add any additional trial. Quality was assessed independently using two distinct tools, the Jadad scale and the Cochrane risk-of-bias tool. The report follows PRISMA, with 265 references identified at the outset. | |
| Volume of available evidence | Very thin |
| FindingFive trials, 264 participants in total. At this scale, every trial carries substantial weight in the conclusion, and a single additional trial could shift the overall picture. | |
| Geographic origin of the trials | All from the same country |
| FindingAll five trials were conducted in China. This does not disqualify them, but it concentrates care practices, concomitant prescribing habits, and recruitment methods, which limits generalization to a European population. | |
| Safety reporting | One trial in five |
| FindingOnly one trial systematically recorded adverse events. The other four report them only partially, headache being the most frequently cited event in four of the five trials. This finding should be read for what it is: an absence of organized recording, not an absence of events. In a minor, incomplete safety reporting is itself grounds for caution. | |
| Risk of bias in included trials | Mixed |
| FindingThe authors do not issue an overall per-trial judgment; they report judgments by domain. Low risk for missing data and for selective reporting across all five trials, low risk for sequence generation and for participant and personnel blinding in four of five trials, but low risk for outcome-assessment blinding in only two of five trials. The authors flag one trial as probably at high risk on this last point. The weighted mean Jadad score is 3.6 (range 3 to 5), classifying all five trials as high quality by this tool’s threshold of 3 or above. On assessor-rated outcomes, the lack of assessor blinding matters more than this classification suggests. | |
| Review protocol registration | Not reported |
| FindingNo prospective registration of the review protocol is reported. This is not a fault, but it is one less safeguard against changes to the analysis plan along the way. | |
| Funding and conflicts of interest | No conflict declared |
| FindingMultiple Chinese public and institutional funding sources, with the authors specifying that the funders had no role in study design, data collection and analysis, the decision to publish, or manuscript preparation. The authors declare no conflict of interest. | |
Results
| Result | What the data show |
|---|---|
| Daily protocol, depressive symptoms | Inconsistent results across trials |
| ReadingTwo trials point in different directions: one reports superiority on the HAMD-24 and the SDS, the other finds no significant difference on the HAMD-17. The authors note that the trial with the more encouraging results is also probably one of the most exposed to bias, while the less exposed trial explicitly reports negative results. This kind of inconsistency does not mean the treatment is ineffective, it means the data cannot settle the question. | |
| Daily protocol, suicidal ideation | Superiority reported in one trial |
| ReadingSuperiority reported on the Chinese version of the Beck Scale for Suicidal Ideation, in a single, single-blind trial, the one the authors judge probably among the most exposed to bias. This is the most encouraging lead in the review, and it rests on a single, unreplicated trial. On an outcome this important, an isolated result should be monitored, not passed on to a patient as a property of the treatment. | |
| Daily protocol, anxiety symptoms | No superiority demonstrated |
| ReadingNo significant difference, in very small samples. Power is insufficient to rule out a modest effect in either direction. | |
| Accelerated protocol, depressive symptoms | Four trials, results the authors judge uncertain |
| ReadingOne trial reports a significant difference on the MADRS, the other three find none on the HAMD-17 or HAMD-24. The accelerated protocol is therefore the one that has been studied the most, four trials against two, without this translating into a consistent result. | |
| Accelerated protocol, other outcomes | No superiority demonstrated over sham |
| ReadingThis covers anxiety in two trials, suicidal ideation in one trial, and cognition in two trials. This is the form people hear about most, and it is the one for which no secondary outcome distinguishes it from sham. Here again, an absence of demonstrated superiority is not a demonstration of absence of effect, especially since the authors themselves describe these trials as underpowered. | |
| Effect sizes | No effect size available in adolescents |
| ReadingAmong their limitations, the authors state that none of the included trials reports an effect size for the primary outcome. The values of 0.38 for response and 0.20 for remission sometimes cited in this context come from a meta-analysis of theta burst stimulation in adults, cited by the authors for comparison, and say nothing about the adolescent population studied in this review. | |
| Tolerability and dropout | Dropout not significantly different, headache more frequent under active stimulation |
| ReadingDropout rates do not differ significantly between groups, which does not establish equivalence given how small the samples are. On tolerability, the result is less neutral than it looks: among the review’s two headline conclusions is that both protocols were associated with a higher incidence of headache, the one trial that recorded this systematically reporting a significantly higher proportion under both the daily and accelerated protocols than under sham. No reassuring conclusion about tolerability can be built on a base where only one trial in five organized the recording. | |
Critical appraisal
| Domain | Judgment |
|---|---|
| Review question and methodology | Sound |
| FindingA clear clinical question, explicit PICOS inclusion criteria, a search run in two languages across six databases, extraction and appraisal carried out independently by three researchers with arbitration from a senior author. The review does its job well, including when that job is to conclude that we do not know. | |
| Evidence base | Insufficient |
| FindingFive trials, 264 participants, protocols highly heterogeneous in target, intensity, and pulse count, ranging from 6,000 to 54,000, with treatment durations from two days to three weeks, and no possible pooling. The quality of the review’s design does not compensate for how little there was to review. | |
| Blinding and outcome assessment | Weak point |
| FindingRisk of bias from lack of outcome-assessor blinding is judged low in only two of five trials, and the authors flag one trial as probably at high risk on this domain. Sham stimulation is also an imperfect comparator by nature, since skin and auditory sensations differ, exposing both participants and assessors to partial unblinding. Two trials are single-blind and one does not specify its blinding arrangements. | |
| Publication bias | Not assessable |
| FindingWith five trials, no asymmetry test is interpretable. A young field with small trials meets the usual conditions for a publication bias favoring positive findings, without this work being able to measure it. | |
| Match between claims and evidence | Exemplary |
| FindingThe authors conclude that evidence is insufficient, write that most trials failed to demonstrate a significant benefit and were underpowered, and call for trials conducted outside China as well as systematic safety recording. No attempt is made to extract a positive signal from material that does not contain one. This is what makes the review useful. | |
| External validity | Limited |
| FindingSingle geographic origin, small samples, two trials including participants up to 24 and 25 years old, and no long-term follow-up data reported. The authors themselves invoke differences in culture, health systems, and diagnostic practices to limit generalization. Transposing these findings to adolescents treated outside China calls for considerable caution. | |
Level of evidence
Confidence is high on one point, and it is a negative one: as things stand, published data do not support the claim that intermittent theta burst stimulation is effective in depressed adolescents. The review looked hard, including in a literature that English-language searches miss, and this is all it found.
Confidence is low on the reverse question. Nothing here demonstrates that the treatment is ineffective. Five trials and 264 participants are not enough to rule out a moderate benefit, and the isolated signal on suicidal ideation justifies continued research.
Confidence is also low on tolerability, for lack of organized recording. Only one trial in five systematically recorded adverse events, and it is precisely that trial that documents a significant increase in headache under active stimulation. In minors, this gap matters as much as the uncertainty about efficacy, and it should shape how any further trial is designed.
The colleague test
What an experienced colleague might say about this study in two minutes, between two consultations.
“Five trials, all from the same country, and only one that systematically records adverse effects in adolescents. And that one finds distinctly more headaches under active stimulation. That’s enough for me to say no when a family shows up with a brochure about the accelerated protocol. I’m not saying it doesn’t work, I’m saying we don’t know, and that’s not the same thing in a minor.”
Translated for practice: the answer to give families is that the available data in adolescents do not support a benefit, and that tolerability has not been documented systematically enough to be presented as reassuring, the one trial that did so reporting more headaches under active stimulation.
What you can do with this
- Do not present intermittent theta burst stimulation as an established treatment for adolescent depression, and do not let a family leave with that impression. The accelerated form is the most studied in this review, four trials out of five, yet it is the one where no secondary outcome distinguishes it from sham.
- Keep the two answers separate when speaking with parents. Efficacy is not established, which is not the same as demonstrated absent. And tolerability is not systematically documented, which is not the same as good, especially since the available signal points to more headaches.
- Reserve the treatment for a research setting, and in that case check that the protocol includes systematic recording of adverse events. In a minor, that is a minimum requirement.
- Check the regulatory framework before referring a patient. The status of this technique in minors, its indications, and the conditions under which it may be used and reimbursed vary from one country to another and are not addressed by this publication; verify them with the competent national authorities at the time of the decision.
- Keep the suicidal-ideation signal as a hypothesis worth following, drawn from a single trial. It should not be passed on to a patient as a property of the treatment.
Frequently asked questions
A parent asks me if it works. What should I tell them?
That the available studies are too few, too small, and too different from one another to draw a conclusion, and that this holds for both the daily and the accelerated forms. This is not a polite way of saying no, it is the actual state of knowledge as the authors themselves describe it.
Is it a problem that all the trials come from the same country?
It is a limit on generalizability, not a judgment on their quality. Associated care practices, concomitant treatments, and recruitment methods differ, which makes extrapolating to adolescents treated outside China uncertain. The authors themselves call for trials conducted elsewhere.
Is the technique dangerous in adolescents?
No serious event is reported in what has been published, but only one trial in five organized the recording of adverse events, which rules out any reassuring conclusion. An absence of recording is not an absence of risk. What is documented shows dropout rates with no significant difference between groups, and a significantly higher incidence of headache under both the daily and accelerated protocols than under sham, in the one trial that measured this.
And what about the finding on suicidal ideation?
It comes from a single trial, under the daily protocol and single-blind, and it has not been replicated. It is also the trial the authors judge probably among the most exposed to bias, and the only trial that assessed this outcome under the accelerated protocol found no difference. On an outcome this important, an isolated result justifies continuing research, not offering the treatment. Presenting it otherwise to a family would be a promise the data do not support.
Why didn’t the authors perform a meta-analysis?
The authors state that meta-analysis was not feasible given the amount of data available. Among their limitations, they also note substantial heterogeneity in stimulation parameters, targets, intensities, and durations, and observe that none of the included trials reports an effect size for the primary outcome. Pooling under these conditions would produce a single number with no clinical meaning.
Annotated bibliography
Source study. Deng YY, Shi ZM, Tang YY, Li Z, Cai DB, Si ZX, Yang XH, Huang XB, Zheng W. “Intermittent theta burst stimulation for adolescent patients with major depressive disorder or bipolar depression: A systematic review.” Journal of Affective Disorders. 2026 Feb 1;394(Pt A):120435. Epub 2025 Oct 14. DOI 10.1016/j.jad.2025.120435. PMID 41101467. The publication includes a note stating that several authors contributed equally, though the extraction available does not allow the exact list of these authors to be established with certainty.
Funding and conflicts of interest. The publication declares multiple Chinese public and institutional funding sources: Guangzhou Science and Technology Program (2023A03J0839, 2023A03J0436, 20251A011047, 20252A011018, 202206010077), Liwan District (Guangzhou) Science and Technology Planning Project (202201012), National Key Clinical Specialty Construction Project [(2023) 33], Guangdong Basic and Applied Basic Research Foundation (2023A1515011383, 2024A1515012578), Chongqing Key Public Health Specialty Construction Project (mental health), Guangzhou Municipal Key Medical Discipline (2021–2023), Guangzhou Science and Technology Plan (2023A03J0827), Guangzhou Traditional Chinese and Integrative Medicine Project (20232A010014), Guangzhou High-Level Key Clinical Specialty, National Key Clinical Specialty Department of Emergency Medicine, Guangzhou Research-Oriented Hospital, and Guangzhou Key Clinical Specialty. The publication states that the funders had no role in study design, data collection and analysis, the decision to publish, or manuscript preparation. Conflict-of-interest statement, verbatim: the authors declare that they have no conflict of interest in conducting this study or in preparing the manuscript, a statement independently confirmed on PubMed. No prospective registration of the review protocol, such as PROSPERO, is mentioned in the text provided. PubMed identifier 41101467 was established on August 31, 2026 by searching bibliographic registries, as the document consulted carried none. Supplementary material was provided and reviewed. It consists of Supplementary Table 1, which gives, for the five trials, all-cause dropout rates and adverse events by group, with proportions and notes flagging a significantly higher incidence of headache under the daily and accelerated protocols than under sham in one trial. This table also indicates that adverse-event data for one trial are not reported. The publication’s two figures, the PRISMA flow diagram and the Cochrane risk-of-bias figure, are images and were not legible in the text provided: the risk-of-bias counts used here therefore come from the text of section 3.3, which states them explicitly. The rightmost columns of Table 1, covering detailed stimulation parameters and trial duration, are truncated in the extraction provided.
Context references. All the references below are cited in the source publication and have not been independently verified. Chu HT, Cheng CM, Liang CS, Chang WH, Juan CH, Huang YZ, Jeng JS, Bai YM, Tsai SJ, Chen MH, Li CT. Efficacy and tolerability of theta-burst stimulation for major depression: a systematic review and meta-analysis. Prog Neuropsychopharmacol Biol Psychiatry. 2021;106:110168. DOI 10.1016/j.pnpbp.2020.110168. Contribution: this is the source of the 0.38 effect size for response and 0.20 for remission. Major limitation: these values concern adults and cannot be transposed to the adolescent population of the review discussed here. Blumberger DM, Vila-Rodriguez F, Thorpe KE, Feffer K, Noda Y, Giacobbe P, Knyahnytska Y, Kennedy SH, Lam RW, Daskalakis ZJ, Downar J. Effectiveness of theta burst versus high-frequency repetitive transcranial magnetic stimulation in patients with depression (THREE-D): a randomised non-inferiority trial. Lancet. 2018;391:1683-1692. DOI 10.1016/S0140-6736(18)30295-2. Contribution: the trial that established theta burst stimulation in adults, showing a shorter session for comparable results. Limitation: adult population, non-inferiority design, no pediatric data. Cole EJ, Phillips AL, Bentzley BS, et al. Stanford neuromodulation therapy (SNT): a double-blind randomized controlled trial. Am J Psychiatry. 2022;179:132-141. DOI 10.1176/appi.ajp.2021.20101429. Contribution: describes the high-dose accelerated protocol with functional-connectivity-guided targeting whose rapid notoriety the review references. Limitation: adult population, small sample, individualized targeting not reproduced in the included adolescent trials. Croarkin PE, Nakonezny PA, Deng ZD, Romanowicz M, Voort JLV, Camsari DD, Schak KM, Port JD, Lewis CP. High-frequency repetitive TMS for suicidal ideation in adolescents with depression. J Affect Disord. 2018;239:282-290. DOI 10.1016/j.jad.2018.06.048. Contribution: a pediatric study of suicidal ideation, cited by the authors as prior work. Limitation: it uses high-frequency repetitive stimulation, not intermittent theta burst stimulation. Zhao Y, He Z, Luo W, Yu Y, Chen J, Cai X, Gao J, Li L, Gao Q, Chen H, Lu F. Effect of intermittent theta burst stimulation on suicidal ideation and depressive symptoms in adolescent depression with suicide attempt: a randomized sham-controlled study. J Affect Disord. 2023;325:618-626. DOI 10.1016/j.jad.2023.01.061. Contribution: the trial from which the signal on suicidal ideation under the daily protocol originates. Limitation: single-blind trial, judged by the review’s authors to be probably among the most exposed to bias. Zhang M, Li W, Ye Y, Hu Z, Zhou Y, Ning Y. Efficacy and safety of intermittent theta burst stimulation on adolescents and young adults with major depressive disorder: a randomized, double blinded, controlled trial. J Affect Disord. 2024;350:214-221. DOI 10.1016/j.jad.2024.01.025. Contribution: the best-protected trial against bias, the only one to systematically record adverse events, and the one documenting the excess of headaches. Limitation: its efficacy results are negative and its population includes young adults up to 25 years old.
Regulatory point. This publication does not address any regulatory framework. The status of intermittent theta burst stimulation in minors, its indications, the conditions under which it may be used, and any coverage or reimbursement vary from one country to another and must be verified, at the time of the decision, with the competent official sources in the relevant jurisdiction, for instance the French authorities for a clinician practicing in France. No claim regarding approval status, indication, or reimbursement can be made from the documents reviewed here.
Editorial collections
Tags
Verified on August 31, 2026 against the full text of the publication and its supplementary material where available. This analysis underwent an independent double reading. The English version was checked for conformity on September 21, 2026, against the figures of the French version and against the source. How we verify what we publish
This analysis is intended for healthcare professionals. It does not constitute a prescribing recommendation and does not replace individual clinical judgment.
Analysis from Psychiatry Evidence Base, evidence-based psychiatry, explained with rigor.